A series
of nanocomposites based on polyamide (NL16, PA) filter
membranes containing metal nanoparticles (NPs) have been prepared
by filtration under reduced pressure of the metal colloids. The ensuing
materials were then investigated as substrates for surface-enhanced
Raman scattering (SERS) imaging studies envisaging the spectroscopic
detection of vestigial organic pollutants dissolved in contaminated
water. The organic dye crystal violet (CV) was used here as a model
pollutant because it is a hazardous compound present in certain effluent
waters. Moreover this compound is well-known for its strong SERS activity,
which is clearly advantageous in the context of material development
for SERS. Indeed, several preparative strategies were employed to
prepare PA-based composites, and the impact on SERS detection was
investigated. These include the use of chemical and morphological
distinct plasmonic NPs (Ag, Au), a variable metal load and changing
the order of addition of the analytical specimens. These studies demonstrate
that the parameters employed in the fabrication of the SERS substrates
have a strong impact on the Raman signal enhancement. The use of Raman
imaging during the fabrication process allows establishing improvements
that translate to better performances of the substrates in the analyte
detection. The results have been interpreted by considering an integrated
set of operational parameters that include the affinity of CV molecules
to the substrate, amount and dispersion of NPs in the PA membranes,
and the detection method. Noteworthy the use of SERS analysis
assisted with Raman imaging allowed achieving a detection limit for
CV as low as 100 aM in ultrapure water and 10 fM in real samples.